Skip to content
Papers.

Chiaroscuro Attention: Spending Compute in the Dark research paper by Accenture, 2026

Accenture · Jun 6, 2026 · Architectures · 0 citations · 1 upvotes · unverified 3 months ago

Read on arXiv

What it shows

CHIAR-Former uses spectral entropy-based routing to dynamically select between DCT, RBF, and self-attention operators, achieving improved efficiency on large text datasets while maintaining performance through hybrid attention mechanisms.

UnverifiedHugging Face's summary; not yet checked by hand.

More from Accenture

All 3
PaperCitations
LayerRoute: Input-Conditioned Adaptive Layer Skipping via LoRA Fine-Tuning for Agentic Language ModelsLayerRoute is a lightweight adapter that selectively skips transformer blocks during inference based on input type, achieving compute savings while maintaining or improving model quality through gated routing and LoRA adaptation.Agents and evaluation · Jun 2026 · Unverified3 months ago1
Memex(RL): Scaling Long-Horizon LLM Agents via Indexed Experience MemoryA memory mechanism called Memex enables large language model agents to handle long-horizon tasks more effectively by maintaining compact context through structured summaries while storing full interaction details in an external database, allowing selective retrieval based on learned criteria.Reasoning · Mar 2026 · Unverified6 months ago12

Other architectures papers

Topic
PaperCitations
Attention Is All You NeedIntroduced the Transformer, the attention-only architecture behind nearly every large language model since.Google · Jun 20179 years ago194k
Mamba: Linear-Time Sequence Modeling with Selective State SpacesA selective state space model that scales linearly with sequence length and matches Transformers on language.Carnegie Mellon University · Dec 20232 years ago9,252
CodeT5: Identifier-aware Unified Pre-trained Encoder-Decoder Models for Code Understanding and GenerationCodeT5, a unified encoder-decoder Transformer model, improves code understanding and generation by leveraging semantic information from identifiers and user comments.Salesforce · Sep 20215 years ago2,510
VGGT: Visual Geometry Grounded TransformerVGGT, a feed-forward neural network, efficiently infers multiple 3D attributes from single or multiple views, outperforming alternatives and enhancing downstream tasks without post-processing.Meta · Mar 2025 · Unverified1 year ago1,807
DeepSeekMoE: Towards Ultimate Expert Specialization in Mixture-of-Experts Language ModelsThe DeepSeekMoE architecture improves expert specialization in Mixture-of-Experts models by segmenting experts and isolating shared ones, achieving better performance and computational efficiency compared to GShard and other models.DeepSeek · Jan 2024 · Unverified2 years ago1,127
CodeT5+: Open Code Large Language Models for Code Understanding and GenerationCodeT5+, a family of flexible encoder-decoder LLMs initialized with frozen LLMs and trained with a mixture of pretraining objectives, achieves state-of-the-art performance across various code-related tasks.Salesforce · May 20233 years ago792
About this paper
Authors
Prateek Kumar Sikdar
arXiv
2606.08327 · PDF
Citations
0, 0 influential · Semantic Scholar
Upvotes
1 · Hugging Face
Lab
Accenture · on Companies · on Acquisitions · on Quarterly · on Paydays

Changes

What changed
Influential citationsfirst count: 0Sep 28, 2026today
Citationsfirst count: 0Sep 28, 2026today
New paperFound by the weekly scan, unverifiedSep 28, 2026today

New papers by email

Monday afternoons, only in weeks with new papers from the labs.

Double opt-in. Unsubscribe any time.